Neural Networks In Brain Research Articles

Bioinspired Nanofluidic Circuits with Integrating Excitatory and Inhibitory Synapses.

Brain neural networks intricately integrate excitatory and inhibitory synaptic potentials to modulate the generation or suppression of action potentials, laying the foundation for neuronal computation. Although bioinspired nanofluidic systems have replicated some synaptic functions, complete integration of postsynaptic potentials remains unachieved. In this work, the developed ion concentration gradient nanofluidic memristor (ICGNM) modulates memristive effects through ion concentration gradient adjustments and exhibits synaptic plasticity phenomena, including paired-pulse facilitation, paired-pulse depression, and spike-rate-dependent plasticity. Furthermore, by incorporation of ICGNMs as the memristive elements into the classic Hodgkin-Huxley model, the action potential generation is replicated. In addition to simulating nanofluidic spiking, these ICGNMs are also employed in a bioinspired nanofluidic circuit to simulate the integration of excitatory and inhibitory synaptic signals, which is highly analogous to the signal integration in actual neural circuits. This work represents a new step toward ionic computing in solution with bioinspired nanofluidic circuits.

Editorial: Symmetry as a guiding principle in artificial and brain neural networks, volume II

Editorial: Symmetry as a guiding principle in artificial and brain neural networks, volume II

Synchronous firing transition between different regions of brain neural networks enhanced by Hamiltonian energy

Synchronous firing transition between different regions of brain neural networks enhanced by Hamiltonian energy

State Feedback control of multiagent singular Linear Systems representing brain neural networks

A multi-agent singular system is an extension of a traditional multi-agent system. The behavior of neural networks within the brain is crucial for cognitive functions, making it essential to understand the learning processes and the development of potential disorders. This study utilizes the analysis of singular linear systems representing brain neural networks to delve into the complexities of the human brain. In this context, the digraph approach is a powerful method for unraveling the intricate neural interconnections. Directed graphs, or digraphs, provide an intuitive visual representation of the causal and influential relationships among different neural units, facilitating a detailed analysis of network dynamics. This work explores the use of digraphs in analyzing singular linear multi-agent systems that model brain neural networks, emphasizing their significance and potential in enhancing our understanding of cognition and brain function.

Spatialspectral-Backdoor: Realizing backdoor attack for deep neural networks in brain–computer interface via EEG characteristics

In recent years, electroencephalogram (EEG) based on the brain–computer interface (BCI) systems have become increasingly advanced, with researcher using deep neural networks as tools to enhance performance. BCI systems heavily rely on EEG signals for effective human–computer interactions, and deep neural networks show excellent performance in processing and classifying these signals. Nevertheless, the vulnerability to backdoor attack is still a major problem. Backdoor attack is the injection of specially designed triggers into the model training process, which can lead to significant security issues. Therefore, in order to simulate the negative impact of backdoor attack and bridge the research gap in the field of BCI, this paper proposes a new backdoor attack method to call researcher attention to the security issues of BCI. In this paper, Spatialspectral-Backdoor is proposed to effectively attack the BCI system. The method is carefully designed to target the spectral active backdoor attack of the BCI system and includes a multi-channel preference method to select the electrode channels sensitive to the target task. Ultimately, the effectiveness of the comparison and ablation experiments is validated on the publicly available BCI competition datasets. The results show that the average attack success rate and clean sample accuracy of Spatialspectral-Backdoor in the BCI scenario are 97.12% and 85.16%, respectively, compared with other backdoor attack methods. Furthermore, by observing the infection ratio of backdoor triggers and visualization of the feature space, the proposed Spatialspectral-Backdoor outperforms other backdoor attack methods.

Resting state functional magnetic resonance imaging in patients with multiple sclerosis before and after high-dose immunosuppressive therapy and autologous hemopoietic stem cell transplantation

BACKGROUND: Multiple sclerosis is a chronic autoimmune disease characterized by multifocal foci of demyelination in the central nervous system, usually affecting people of working age. The disease causes damage to the blood-brain barrier, the development of multifocal inflammation, destruction of the myelin sheath of axons and various degrees of damage. It is clinically manifested by restriction of motor activity, visual acuity, as well as other symptoms leading to loss of performance and disability of the patient. AIM: determination of changes in the functional connectivity of brain neural networks in patients with multiple sclerosis before and after high-dose immunosuppressive therapy and autologous hematopoietic stem cell transplantation by performing functional magnetic resonance imaging at rest. MATERIALS AND METHODS: The data of functional magnetic resonance imaging of patients with multiple sclerosis were analyzed in dynamics before and after the use of high-dose immunosuppressive therapy followed by autologous hematopoietic stem cell transplantation. The study involved 25 patients with a verified diagnosis of multiple sclerosis. Each underwent complex magnetic resonance imaging at two time points (before and after high-dose immunosuppressive therapy followed by autologous hematopoietic stem cell transplantation) with a difference of 12 months, which included structural magnetic resonance imaging - in order to exclude the presence of pathological foci in the brain (in addition to foci of multiple sclerosis) and functional magnetic resonance imaging.-resonance imaging at rest — to assess functional connectivity. Also, according to the method generally accepted in classical neurology, a clinical neurological examination was performed. RESULTS: At the stage of comparing data on the two groups obtained using functional magnetic resonance imaging at rest, changes in functional activity were detected in various parts of the brain, presumably responsible for clinical differences in the studied groups. CONCLUSION: Currently, the links between brain structures and morphological changes that cause cognitive impairment in multiple sclerosis are being studied. To predict the progression of the disease, the development of biomarkers, including those based on functional magnetic resonance imaging, is required. Evaluating changes in the functional connectivity of brain neural networks can help personalize therapeutic and rehabilitation approaches.

Effective Emotion Recognition by Learning Discriminative Graph Topologies in EEG Brain Networks.

Multichannel electroencephalogram (EEG) is an array signal that represents brain neural networks and can be applied to characterize information propagation patterns for different emotional states. To reveal these inherent spatial graph features and increase the stability of emotion recognition, we propose an effective emotion recognition model that performs multicategory emotion recognition with multiple emotion-related spatial network topology patterns (MESNPs) by learning discriminative graph topologies in EEG brain networks. To evaluate the performance of our proposed MESNP model, we conducted single-subject and multisubject four-class classification experiments on two public datasets, MAHNOB-HCI and DEAP. Compared with existing feature extraction methods, the MESNP model significantly enhances the multiclass emotional classification performance in the single-subject and multisubject conditions. To evaluate the online version of the proposed MESNP model, we designed an online emotion monitoring system. We recruited 14 participants to conduct the online emotion decoding experiments. The average online experimental accuracy of the 14 participants was 84.56%, indicating that our model can be applied in affective brain-computer interface (aBCI) systems. The offline and online experimental results demonstrate that the proposed MESNP model effectively captures discriminative graph topology patterns and significantly improves emotion classification performance. Moreover, the proposed MESNP model provides a new scheme for extracting features from strongly coupled array signals.

Modeling short visual events through the BOLD moments video fMRI dataset and metadata

Studying the neural basis of human dynamic visual perception requires extensive experimental data to evaluate the large swathes of functionally diverse brain neural networks driven by perceiving visual events. Here, we introduce the BOLD Moments Dataset (BMD), a repository of whole-brain fMRI responses to over 1000 short (3 s) naturalistic video clips of visual events across ten human subjects. We use the videos’ extensive metadata to show how the brain represents word- and sentence-level descriptions of visual events and identify correlates of video memorability scores extending into the parietal cortex. Furthermore, we reveal a match in hierarchical processing between cortical regions of interest and video-computable deep neural networks, and we showcase that BMD successfully captures temporal dynamics of visual events at second resolution. With its rich metadata, BMD offers new perspectives and accelerates research on the human brain basis of visual event perception.

Representations and generalization in artificial and brain neural networks

Humans and animals excel at generalizing from limited data, a capability yet to be fully replicated in artificial intelligence. This perspective investigates generalization in biological and artificial deep neural networks (DNNs), in both in-distribution and out-of-distribution contexts. We introduce two hypotheses: First, the geometric properties of the neural manifolds associated with discrete cognitive entities, such as objects, words, and concepts, are powerful order parameters. They link the neural substrate to the generalization capabilities and provide a unified methodology bridging gaps between neuroscience, machine learning, and cognitive science. We overview recent progress in studying the geometry of neural manifolds, particularly in visual object recognition, and discuss theories connecting manifold dimension and radius to generalization capacity. Second, we suggest that the theory of learning in wide DNNs, especially in the thermodynamic limit, provides mechanistic insights into the learning processes generating desired neural representational geometries and generalization. This includes the role of weight norm regularization, network architecture, and hyper-parameters. We will explore recent advances in this theory and ongoing challenges. We also discuss the dynamics of learning and its relevance to the issue of representational drift in the brain.

Disrupted individual-level morphological brain network in spinal muscular atrophy types 2 and 3.

Spinal muscular atrophy (SMA) is one of the most common monogenic neuromuscular diseases, and the pathogenesis mechanisms, especially the brain network topological properties, remain unknown. This study aimed to use individual-level morphological brain network analysis to explore the brain neural network mechanisms in SMA. Individual-level gray matter (GM) networks were constructed by estimating the interregional similarity of GM volume distribution using both Kullback-Leibler divergence-based similarity (KLDs) and Jesen-Shannon divergence-based similarity (JSDs) measurements based on Automated Anatomical Labeling 116 and Hammersmith 83 atlases for 38 individuals with SMA types 2 and 3 and 38 age- and sex-matched healthy controls (HCs). The topological properties were analyzed by the graph theory approach and compared between groups by a nonparametric permutation test. Additionally, correlation analysis was used to assess the associations between altered topological metrics and clinical characteristics. Compared with HCs, although global network topology remained preserved in individuals with SMA, brain regions with altered nodal properties mainly involved the right olfactory gyrus, right insula, bilateral parahippocampal gyrus, right amygdala, right thalamus, left superior temporal gyrus, left cerebellar lobule IV-V, bilateral cerebellar lobule VI, right cerebellar lobule VII, and vermis VII and IX. Further correlation analysis showed that the nodal degree of the right cerebellar lobule VII was positively correlated with the disease duration, and the right amygdala was negatively correlated with the Hammersmith Functional Motor Scale Expanded (HFMSE) scores. Our findings demonstrated that topological reorganization may prioritize global properties over nodal properties, and disrupted topological properties in the cortical-limbic-cerebellum circuit in SMA may help to further understand the network pathogenesis underlying SMA.

Repurposing Ketamine in the Therapy of Depression and Depression-Related Disorders: Recent Advances and Future Potential.

Depression represents a prevalent and enduring mental disorder of significant concern within the clinical domain. Extensive research indicates that depression is very complex, with many interconnected pathways involved. Most research related to depression focuses on monoamines, neurotrophic factors, the hypothalamic-pituitary-adrenal axis, tryptophan metabolism, energy metabolism, mitochondrial function, the gut-brain axis, glial cell-mediated inflammation, myelination, homeostasis, and brain neural networks. However, recently, Ketamine, an ionotropic N-methyl-D-aspartate (NMDA) receptor antagonist, has been discovered to have rapid antidepressant effects in patients, leading to novel and successful treatment approaches for mood disorders. This review aims to summarize the latest findings and insights into various signaling pathways and systems observed in depression patients and animal models, providing a more comprehensive view of the neurobiology of anxious-depressive-like behavior. Specifically, it highlights the key mechanisms of ketamine as a rapid-acting antidepressant, aiming to enhance the treatment of neuropsychiatric disorders. Moreover, we discuss the potential of ketamine as a prophylactic or therapeutic intervention for stress-related psychiatric disorders.

Introduction to the assessment and management of persistent postural-perceptual dizziness.

Persistent postural-perceptual dizziness classifies patients with chronic dizziness, often triggered by an acute episode of vestibular dysfunction or threat to balance. Unsteadiness and spatial disorientation vary in intensity but persist for over three months, exacerbated by complex visual environments. Literature suggests diagnosis relies on a clinical history of persistent subjective dizziness and normal vestibular and neurological examination findings. Behavioural diagnostic biomarkers have been proposed, to facilitate diagnosis. Research has focused on understanding the neural mechanisms that underpin this perceptual disorder, with imaging data supporting altered connectivity between neural brain networks that process vision, motion and emotion. Behavioural research identified the perceptual and motor responses to a heightened perception of imbalance. Management utilises head and body motion detection, and downregulation of visual motion excitability, reducing postural hypervigilance and anxiety. Combinations of physical and cognitive therapies, with antidepressant medications, help if the condition is associated with mood disorder.

Percolation transitions in partially edge-coupled interdependent networks with different group size distributions

In many systems, from brain neural networks to epidemic transmission networks, pairwise interactions are insufficient to express complex relationships. Nodes sometimes cooperate and form groups to increase their robustness to risks, and each such group can be considered a “supernode”. Furthermore, previous studies of cascading failures in interdependent networks have typically concentrated on node coupling connections; however, in many realistic scenarios, interactions occur between the edges connecting nodes rather than between the nodes themselves. Networks of this type are called edge-coupled interdependent networks. To better reflect complex networks in the real world, in this paper, we construct a theoretical model of a two-layer partially edge-coupled interdependent network with groups, where all nodes in the same group are functionally dependent on each other. We identify several types of phase transitions, namely, discontinuous, hybrid and continuous, which are determined by the strength of the dependency and the distribution of the supernodes. We first apply our developed mathematical framework to ErdsRnyi and scale-free partially edge-coupled interdependent networks with equally sized groups to analytically and numerically calculate the phase transition thresholds and the critical dependency strengths that distinguish different types of transitions. We then investigate the influence of the group size distribution on cascading failures by presenting examples of two different heterogeneous group size distributions. Our theoretical predictions and numerical findings are in close agreement, demonstrating that decreasing the dependency strength and increasing group size heterogeneity can increase the robustness of interdependent networks. Our results have significant implications for the design and optimization of network security and fill a knowledge gap in the robustness of partially edge-coupled interdependent networks with different group size distributions.

Neural Mechanisms of Nonauditory Effects of Noise Exposure on Special Populations.

Due to the abnormal structure and function of brain neural networks in special populations, such as children, elderly individuals, and individuals with mental disorders, noise exposure is more likely to have negative psychological and cognitive nonauditory effects on these individuals. There are unique and complex neural mechanisms underlying this phenomenon. For individuals with mental disorders, there are anomalies such as structural atrophy and decreased functional activation in brain regions involved in emotion and cognitive processing, such as the prefrontal cortex (PFC). Noise exposure can worsen these abnormalities in relevant brain regions, further damaging neural plasticity and disrupting normal connections and the transmission of information between the PFC and other brain areas by causing neurotransmitter imbalances. In the case of children, in a noisy environment, brain regions such as the left inferior frontal gyrus and PFC, which are involved in growth and development, are more susceptible to structural and functional changes, leading to neurodegenerative alterations. Furthermore, noise exposure can interrupt auditory processing neural pathways or impair inhibitory functions, thus hindering children's ability to map sound to meaning in neural processes. For elderly people, age-related shrinkage of brain regions such as the PFC, as well as deficiencies in hormone, neurotransmitter, and nutrient levels, weakens their ability to cope with noise. Currently, it is feasible to propose and apply coping strategies to improve the nonauditory effects of noise exposure on special populations based on the plasticity of the human brain.

Neurocomputational mechanisms underlying perception and sentience in the neocortex.

The basis for computation in the brain is the quantum threshold of "soliton," which accompanies the ion changes of the action potential, and the refractory membrane at convergences. Here, we provide a logical explanation from the action potential to a neuronal model of the coding and computation of the retina. We also explain how the visual cortex operates through quantum-phase processing. In the small-world network, parallel frequencies collide into definable patterns of distinct objects. Elsewhere, we have shown how many sensory cells are meanly sampled from a single neuron and that convergences of neurons are common. We also demonstrate, using the threshold and refractory period of a quantum-phase pulse, that action potentials diffract across a neural network due to the annulment of parallel collisions in the phase ternary computation (PTC). Thus, PTC applied to neuron convergences results in a collective mean sampled frequency and is the only mathematical solution within the constraints of the brain neural networks (BNN). In the retina and other sensory areas, we discuss how this information is initially coded and then understood in terms of network abstracts within the lateral geniculate nucleus (LGN) and visual cortex. First, by defining neural patterning within a neural network, and then in terms of contextual networks, we demonstrate that the output of frequencies from the visual cortex contains information amounting to abstract representations of objects in increasing detail. We show that nerve tracts from the LGN provide time synchronization to the neocortex (defined as the location of the combination of connections of the visual cortex, motor cortex, auditory cortex, etc.). The full image is therefore combined in the neocortex with other sensory modalities so that it receives information about the object from the eye and all the abstracts that make up the object. Spatial patterns in the visual cortex are formed from individual patterns illuminating the retina, and memory is encoded by reverberatory loops of computational action potentials (CAPs). We demonstrate that a similar process of PTC may take place in the cochlea and associated ganglia, as well as ascending information from the spinal cord, and that this function should be considered universal where convergences of neurons occur.

Brain-like illusion produced by Skye's Oblique Grating in deep neural networks.

The analogy between the brain and deep neural networks (DNNs) has sparked interest in neuroscience. Although DNNs have limitations, they remain valuable for modeling specific brain characteristics. This study used Skye's Oblique Grating illusion to assess DNNs' relevance to brain neural networks. We collected data on human perceptual responses to a series of visual illusions. This data was then used to assess how DNN responses to these illusions paralleled or differed from human behavior. We performed two analyses:(1) We trained DNNs to perform horizontal vs. non-horizontal classification on images with bars tilted different degrees (non-illusory images) and tested them on images with horizontal bars with different illusory strengths measured by human behavior (illusory images), finding that DNNs showed human-like illusions; (2) We performed representational similarity analysis to assess whether illusory representation existed in different layers within DNNs, finding that DNNs showed illusion-like responses to illusory images. The representational similarity between real tilted images and illusory images was calculated, which showed the highest values in the early layers and decreased layer-by-layer. Our findings suggest that DNNs could serve as potential models for explaining the mechanism of visual illusions in human brain, particularly those that may originate in early visual areas like the primary visual cortex (V1). While promising, further research is necessary to understand the nuanced differences between DNNs and human visual pathways.

A Cross-Sectional and Longitudinal Integrated Study on Brain Functional Changes in a Neuropathic Pain Rat Model.

Human and animal imaging studies demonstrated that chronic pain profoundly alters the structure and the functionality of several brain regions and even causes mental dysfunctions such as depression and anxiety disorders. In this article, we conducted a multimodal study cross-sectionally and longitudinally, to evaluate how neuropathic pain affects the brain. Using the spared nerve injury (SNI) model which promotes long-lasting mechanical allodynia, results showed that neuropathic pain deeply modified the intrinsic organization of the brain functional network 2 weeks after injury. There are significant changes in the activity of the left thalamus (Th_L) and left olfactory bulb (OB_L) brain regions after SNI, as evidenced by both the blood oxygen level-dependent (BOLD) signal and c-Fos expression. Importantly, these changes were closely related to mechanical pain behavior of rats. However, it is worth noting that after morphine administration for analgesia, only the increased activity in the TH region is reversed, while the decreased activity in the OB region becomes more prominent. Functional connectivity (FC) and c-Fos correlation analysis further showed these two regions of interest (ROIs) exhibit different FC patterns with other brain regions. Our study comprehensively revealed the adaptive changes of brain neural networks induced by nerve injury in both cross-sectional and longitudinal dimensions and emphasized the abnormal activity and FC of Th_L and OB_L in the pathological condition. It provides reliable assistance in exploring the intricate mechanisms of diseases.

CyberWatch: Deep Learning-Driven Network Intrusion Detection

Abstract—The present era is being dominated by the digital world which has also given rise to growing cyber threats and crimes, the project introduces an innovative intrusion detection system(IDS) that deploys deep learning’s pattern recognition capabilities with a real-time algorithm approach to threats and network security. The project’s framework involves two key components: Graph theory and Artificial Neural Network (ANN). The graph theory is used to represent the IoT network structure which helps represent relations between network structures and analyze the anomalies or malicious activities present in the network. Artificial Neural Network (ANN) model, on the other hand, is based on a human’s brain neural network; which works as a machine learning algorithm to recognize and predict patterns. Cyberwatch is trained on various datasets such as CIC IDS 2018 and CIC IDS 2017. The project’s main aim is to provide an effective solution to the ever-changing landscape of network intrusion. Keywords—Intrusion Detection System(IDS), Energy Ef- ficiency, Internet Of Things(IoT), Network Forensics, Graph Theory Support Vector Machine(SVM), Genetic Algorithm, Artificial Neural Networks(ANN), Cybercrime

Modeling for neurosurgical laser interstitial thermal therapy with and without intracranial recording electrodes

Laser thermal ablation has become a prominent neurosurgical treatment approach, but in epilepsy patients it cannot currently be safely implemented with intracranial recording electrodes that are used to study interictal or epileptiform activity. There is a pressing need for computational models of laser interstitial thermal therapy (LITT) with and without intracranial electrodes to enhance the efficacy and safety of optical neurotherapies. In this paper, we aimed to build a biophysical bioheat and ray optics model to study the effects of laser heating in the brain, with and without intracranial electrodes in the vicinity of the ablation zone during the LITT procedure. COMSOL Multiphysics finite element method (FEM) solver software was used to create a bioheat thermal model of brain tissue, with and without blood flow incorporation via Penne's model, to model neural tissue response to laser heating. We report that the close placement of intracranial electrodes can increase the maximum temperature of the brain tissue volume as well as impact the necrosis region volume if the electrodes are placed too closely to the laser coupled diffuse fiber tip. The model shows that an electrode displacement of 4 mm could be considered a safe distance of intracranial electrode placement away from the LITT probe treatment area. This work, for the first time, models the impact of intracranially implanted recording electrodes during LITT, which could improve the understanding of the LITT treatment procedure on the brain's neural networks a sufficient safe distance to the implanted intracranial recording electrodes. We recommend modeling safe distances for placing the electrodes with respect to the infrared laser coupled diffuse fiber tip.

Adaptive Synaptic Scaling in Spiking Networks for Continual Learning and Enhanced Robustness.

Synaptic plasticity plays a critical role in the expression power of brain neural networks. Among diverse plasticity rules, synaptic scaling presents indispensable effects on homeostasis maintenance and synaptic strength regulation. In the current modeling of brain-inspired spiking neural networks (SNN), backpropagation through time is widely adopted because it can achieve high performance using a small number of time steps. Nevertheless, the synaptic scaling mechanism has not yet been well touched. In this work, we propose an experience-dependent adaptive synaptic scaling mechanism (AS-SNN) for spiking neural networks. The learning process has two stages: First, in the forward path, adaptive short-term potentiation or depression is triggered for each synapse according to afferent stimuli intensity accumulated by presynaptic historical neural activities. Second, in the backward path, long-term consolidation is executed through gradient signals regulated by the corresponding scaling factor. This mechanism shapes the pattern selectivity of synapses and the information transfer they mediate. We theoretically prove that the proposed adaptive synaptic scaling function follows a contraction map and finally converges to an expected fixed point, in accordance with state-of-the-art results in three tasks on perturbation resistance, continual learning, and graph learning. Specifically, for the perturbation resistance and continual learning tasks, our approach improves the accuracy on the N-MNIST benchmark over the baseline by 44% and 25%, respectively. An expected firing rate callback and sparse coding can be observed in graph learning. Extensive experiments on ablation study and cost evaluation evidence the effectiveness and efficiency of our nonparametric adaptive scaling method, which demonstrates the great potential of SNN in continual learning and robust learning.